Exploring the drivers of XAI-enhanced clinical decision support systems adoption: Insights from a stimulus-organism-response perspective

The concept of Explainable Artificial Intelligence (XAI) provides a clear and comprehensible explanation for the reasoning behind a system's output, allowing users to understand the context in which it operates. In the realm of Clinical Decision Support Systems (CDSS), XAI is particularly crucial, as it helps healthcare professionals (HCPs) in their decision-making processes. Without XAI, there is a risk of over-reliance on the system's output, poten-tially leading to subpar results. Despite the numerous benefits that XAI-enhanced CDSS hold in the healthcare industry, there has been a limited number of studies examining their implementation and acceptance. Thus, the objective of this study was to examine the adoption of XAI-based CDSS through the Stimulus-Organism-Response model. The sample consisted of 172 HCPs from Malaysian public and private hospitals, and the research model was tested using Partial Least Squares Structural Equation Modeling (PLS-SEM) in conjunction with the boot-strapping method. The results showed a significant positive correlation between the stimulus factors of informed action, transparent interaction, and representational fidelity and the positive attitude towards XAI-based CDSS as an organism factor. Additionally, the study found that attitude was a significant predictor of the intention to adopt as a response factor. The analysis also revealed a negative and significant moderation effect of perceived performance risk on the relationship between attitude and intention, while the positive moderating effect of perceived fairness was not supported. The findings of this study have significant implications for both theoretical and practical considerations and highlight the importance of XAI in the field of Clinical Decision Support Systems.

Automated summary

The concept of Explainable Artificial Intelligence (XAI) provides a clear and comprehensible explanation for system outputs. It is particularly crucial in Clinical Decision Support Systems (CDSS) as it helps healthcare professionals in decision-making. However, limited studies have examined the implementation and acceptance of XAI-enhanced CDSS. This study aims to examine XAI adoption through the Stimulus-Organism-Response model, using a sample of 172 HCPs from Malaysian hospitals. The findings highlight the importance of XAI in CDSS and have significant implications for theory and practice.